Noise Susceptibility Studies / Magnetic Field Tests - Status & Plans of the Aachen Group

Size: px

Start display at page:

Download "Noise Susceptibility Studies / Magnetic Field Tests - Status & Plans of the Aachen Group"

Archibald Price
5 years ago
Views:

1 Noise Susceptibility Studies / Magnetic Field Tests - Status & Plans of the Aachen Group Lutz Feld, Rüdiger Jussen, Waclaw Karpinski, Katja Klein, Jennifer Merz, Jan Sammet 1. Physikalisches Institut B, RWTH Aachen University Tracker Upgrade Power WG Meeting June 4 th, 2009

2 Outline Personal & funding Noise susceptibility studies Magnetic field test of DC-DC converters Plans Summary 2

Update on Personal Lutz Feld: team leader Waclaw Karpinski: electronics engineer plus electronics workshop team Katja Klein: Helmholtz Alliance fellow

3 Update on Personal Lutz Feld: team leader Waclaw Karpinski: electronics engineer plus electronics workshop team Katja Klein: Helmholtz Alliance fellow (4-years from April 08) Two PhD students: Jan Sammet Rüdiger Jussen Diploma student: Jennifer Merz (Effect of powering schemes on the material budget) 3

4 Funding News We have received from BMBF (= main German funding body for HE physics) for the next 3-year funding period, starting July 09: invest money 3 PhD positions for SLHC 4

5 Summary of Activities Investigation of system aspects of novel powering schemes PCB development & system tests separate talk today by Jan Sammet Noise susceptibility measurements Noise injection into silicon strip modules covered in this talk Contribute to the development & characterization of magnetic field tolerant and radiation hard DC-DC buck converters, in coll. with CERN PH-ESE group Magnetic field test covered in this talk Integration and test of CERN converters with CMS strip modules only short summary of plans today Simulation of material budget of different powering schemes final results ready, will be presented in the next meeting 5

Noise Susceptibility Studies Goal: identify particularly critical bandwidth(s) for converter switching frequency Bulk current injection (BCI) test-stand has been

6 Noise Susceptibility Studies Goal: identify particularly critical bandwidth(s) for converter switching frequency Bulk current injection (BCI) test-stand has been set up (Rüdiger Jussen) A noise current of 70dB A (I eff = 3.16mA) is injected into the power lines Differential Mode (DM) and Common Mode (CM) on 2.5V and 1.25V 6

LISN Injection & current probe BCI Set-up Current probe in CM

Amplifier Spectrum analyzer Noise injection into one module

7 LISN Injection & current probe BCI Set-up Current probe in CM configuration Power supplies Petal Frequency generator Amplifier Spectrum analyzer Noise injection into one module (6.4) Noise is injected into a single module Frequency swept from 100kHz 100MHz Step width: 0.1MHz between 100kHz and 10MHz, 1.0MHz between 10MHz and 30MHz, 2.5MHz between 30MHz and 100MHz 7

8 Effects on Module Noise 8

9 Effects on Module Noise Mean noise of APV2 Noise of strip 512 Edge strips much more sensitive due to their coupling to the bias ring On-chip common mode subtraction is very efficient for most strips More in back-up slides Concentrate on edge strips 9

10 Peak Mode Peak at 6-8MHz, not at 1/(2 50ns) = 3.2MHz, as expected from shaping time Higher susceptibility for differential mode and 1.25V = pre-amp reference voltage Peak position independent of injected amplitude or module position 10

11 Peak vs. Deconvolution Mode Deconvolution mode Peak mode Slight shift of peaks Interpretation difficult 11

12 What about Higher Frequencies? Zoom Cable resonances can be observed if cable length L = n /4 Two open ends (LISN 50, module ~ 2 ) L = /2 Cable length varied between ~ 1.1m, 1.5m, 2.1m f = 89.8MHz, 65.9MHz, 47.0MHz Measurements above ~ 30MHz are not reliable But no shift of peaks below 30MHz 12

13 Influence of Pre-amp Reference Voltage APV25 pre-amplifier DM, 2.5V strip V125 V250 bias ring VSS=GND [Mark Raymond] [Hybrid] connected to to Ground V125 Edge strips are capacitively coupled to bias ring Bias ring referenced to ground, pre-amp to 1.25V Bias ring connected to 1.25V instead of ground Susceptibility decreases drastically Pre-amp should be referenced to ground 13

14 BCI Summary Results are ~ consistent with measurements of Fernando Arteche (2004) Powerful method, but interpretation difficult (needs modelling) Shorter measurement time needed for faster turnaround (now ~ 1d per curve) automation of measurement with LabView is foreseen (Rüdiger) Will be useful to characterize susceptibility of SLHC devices (hybrids, modules,...) F. Arteche, measurements with TEC petal in Aachen, 2004 (SLAC-PUB-11886, May 2006 ) 14

Magnet Test DC-DC converters must function in ~ 4T magnetic field no magnetic components Tests with 7T NMR-magnet at Forschungszentrum Jülich, close to Aachen Enpirion and

15 Magnet Test DC-DC converters must function in ~ 4T magnetic field no magnetic components Tests with 7T NMR-magnet at Forschungszentrum Jülich, close to Aachen Enpirion and CERN AMIS1 buck converters + LBNL charge pump tested (Rüdiger) both versions with air-core and ferrite coils 13 DC-DC converters tested in total (can show only examples here) 15

16 Magnet Test Set-up Handle for probes B field 9m long BNC cables Handle being inserted into magnet Sourcemeter = Load Scope with probes Magnet PS Windows-PC running LabView 16

17 Efficiency Measurement Regulated by converter Value set in sourcemeter and monitored with current probe Eff Vout Iout V I R I 2 in in in Set and measured with PS Measured with PS Correction for cable losses Note: the output voltage was not measured this must be changed in the future! Efficiency was measured inside and outside of magnet with same set-up Measurement of other observables (ripple?) difficult due to long cables what else should be measured? 17

18 Efficiency (7T) / Efficiency (0T) Enpirion EQ5382D Buck Converter Enpirion with ferrite coil V out = 1.25V Severe efficiency loss with ferrite inductor Efficiency change < 0.5% with air-core inductor Enpirion with air-core toroid V out = 1.25V Eff. (7T) / Eff. (0T) 18

19 AMIS1 Buck Converter w/ Air-Core Solenoid Efficiency (7T) / Efficiency (0T) Efficiency changes by less than 5% with air-core inductor Reason for larger deviations wrt Enpirion not clear, converter stability? With ferrite inductor, PS went in over-current condition (back-up slides) 19

20 LBNL Charge Pump No efficiency change for V out = 2.5V For V out = 1.25V, converter was probably not in same state (stability problems) 20

21 Magnet Test Summary No surprises: All converters with ferrite coils showed severe efficiency loss or over-current All converters with air-core inductors, plus charge pumps, worked without significant efficiency loss We know now how to do the measurements and what to improve Measure output voltage Test various coil orientations Time needed for a measurement campaign: 2 days (but must be arranged) Suggest to repeat test with CERN ASIC in IHP technology and improved set-up maybe also with AMIS2? 21

22 Future Plans System test with strip modules of CERN buck converter PCB with discrete components (started) CERN AMIS1 with Bristol PCB inductors (asap) CERN AMIS2 buck converter ASIC (summer) CERN IHP buck converter ASIC (autumn) PCB development for integration of DC-DC converters into tracker structures Automate and improve several existing test-stands (BCI, EMI, efficiency) Set up EMI-scanner to investigate coupling mechanisms of radiated noise Continue material budget studies Develop specifications for buck converter Get more practical experience with charge pumps 22

23 Summary & Conclusions A bulk current injection test-bench for noise susceptibility studies has been set-up and first measurements have been performed Various DC-DC converters have been tested in a 7T magnetic field Both set-ups need some improvements, but seem to be useful for tests of future converter and module prototypes Simulation of material budget for powering/cooling schemes finished, will be presented in the next meeting 23

24 Back-up Slides 24

25 The APV25 f = 1/(2 50nsec) = 3.2MHz 25

26 The APV V 2.5V * * is connected to 2.5V since about

27 The APV V 2.5V 27

28 On-Chip Common Mode Subtraction 128 APV inverter stages powered from 2.5V via common resistor (historical reasons) mean common mode (CM) of all 128 channels is effectively subtracted on-chip Works fine for regular channels which see mean CM CM appears on open channels which see less CM than regular channels CM imperfectly subtracted for channels with increased noise, i.e. edge channels strip pre-amplifier V125 V250 v IN +v CM inverter V250 R (external) v CM v OUT = -v IN VSS Node is common to all 128 inverters in chip 28

29 Common Mode & Differential Mode Differential Mode (DM): Source Load Common Mode (CM): Source Load 29

30 AMIS1 Buck Converter w/ Ferrite Coil 30

DC-DC Conversion Studies Status Report from Aachen

DC-DC Conversion Studies Status Report from Aachen Lutz Feld, Rüdiger Jussen, Waclaw Karpinski, Katja Klein, Jennifer Merz, Jan Sammet 1. Physikalisches Institut B, RWTH Aachen University Tracker Upgrade